Shared customer verification network
A common database of documented incidents that independent businesses both contribute to and check against, so that a customer’s behaviour at one company is visible to the next before a contract is signed. Its defining property is reciprocity: value comes from how many independent members report, not from how many subscribe. Full definition.
Internal blacklist
A private list of customers a single company refuses to serve. It works, and it is blind past the edge of that company – it catches a person the second time they come to you, which for someone who never returns to the same counter is never. How the two differ.
Blind index
A way of searching a database for a value without storing the value. The identifier is turned into an irreversible cryptographic fingerprint (HMAC-SHA256 with a secret key) before it is sent; two identical inputs produce the same fingerprint, so a match can be confirmed, but the original cannot be reconstructed from it – not by an attacker who steals the database, and not by us.
Risk traffic light
A five-state summary of what the network holds about an identifier: none, positive, low, medium or high. It reports and never decides. Green means no record was found, which is not a guarantee – only that the network has not seen this person before.
Decision support
A system that supplies facts to a person who then decides, as opposed to one that decides on its own. The distinction is legal, not stylistic: Article 22 of the GDPR prohibits decisions based solely on automated processing where they significantly affect a person, so a verification result must never arrive as an instruction.
Evidence requirement
The rule that no incident can be recorded without at least one documented item – a signed handover protocol, photographs, an unpaid invoice, a damage or police report, a booking-platform record or a witness statement. It is enforced by the form itself, and it is the line between a documented incident and defamation.
Incident
A documented event in which a customer failed to pay, damaged something, did not return it, misused a contract or acted abusively. It is a record of what happened, with evidence attached – not an opinion about the person, and not a judgement about future behaviour.
Rental fraud prevention
The practice of recognising, before a contract is signed, a customer who intends not to pay or not to return. It differs from ordinary risk management because the same person repeats the pattern across companies rather than within one, so prevention that relies only on your own history is structurally late. More on this.
Guest screening
Checking a guest before a stay in accommodation – short-term rentals, apartments, hotels. The risks are specific to the setting: parties, unregistered extra guests, damage to the property and bookings made under someone else’s identity. More on this.
Tenant verification
Checking a prospective tenant before a lease. Where a credit check asks whether someone can pay, tenant verification in a shared network asks what happened at previous rentals – unpaid months, damage beyond wear, or a departure without handover. More on this.
Customer verification platform
Software that checks a customer against a source of risk information at the moment of a transaction. What distinguishes platforms in this category is the source: an internal list, a public register, or a network shared across independent companies. More on this.
Data subject
The person a record is about, in the language of the GDPR. A data subject has the right to know that a record exists, to see what it contains, to challenge it as inaccurate and to have it erased – and those rights are exercised against the network, not against the company that filed the record.
Audit log
An immutable record of who accessed what and when. Immutable is meant literally: the database permits inserts and refuses updates and deletions, including from administrators, because a log that can be edited proves nothing about the period before it was edited.
Network effect
The property that each new member makes the system more valuable to every existing one. In a verification network it is the whole economics: software can be rewritten in a quarter, but the density of shared memory is what a competitor cannot copy – and it also means an early network is honestly worth less than a mature one.